1-s2.0-S092325080kjghkhj6001264-main

8/17/2019 1-s2.0-S092325080kjghkhj6001264-main

1/9

Research in Microbiology 157 (2006) 867–875www.elsevier.com/locate/resmic

Effects of nutritional and environmental conditionson Sinorhizobium meliloti biolm formation

Luciana Rinaudi a, Nancy A. Fujishige b, Ann M. Hirsch b,c, Erika Banchio a,Angeles Zorreguieta d, Walter Giordano a,∗

a Departamento de Biología Molecular, Universidad Nacional de Río Cuarto, 5800-Río Cuarto, Córdoba, Argentinab Department of Molecular, Cell and Developmental Biology, University of California-Los Angeles, 405 Hilgard Avenue, Los Angeles, CA 90095-1606, USA

c Molecular Biology Institute, University of California-Los Angeles, 405 Hilgard Avenue, Los Angeles, CA 90095-1606, USAd Fundación Instituto Leloir, CONICET and Instituto de Investigaciones Bioquímicas, FCEyN, Universidad de Buenos Aires, Patricias Argentinas 435,

C1405BWE-Buenos Aires, Argentina

Received 15 February 2006; accepted 8 June 2006

Available online 12 July 2006

Abstract

Rhizobia are non-spore-forming soil bacteria that x atmospheric nitrogen into ammonia in a symbiosis with legume roots. However, in theabsence of a legume host, rhizobia manage to survive and hence must have evolved strategies to adapt to diverse environmental conditions. Thecapacity to respond to variations in nutrient availability enables the persistence of rhizobial species in soil, and consequently improves theirability to colonize and to survive in the host plant. Rhizobia, like many other soil bacteria, persist in nature most likely in sessile communitiesknown as biolms, which are most often composed of multiple microbial species. We have been employing in vitro assays to study environmentalparameters that might inuence biolm formation in the Medicago symbiont Sinorhizobium meliloti . These parameters include carbon source,

amount of nitrate, phosphate, calcium and magnesium as well as the effects of osmolarity and pH. The microtiter plate assay facilitates thedetection of subtle differences in rhizobial biolms in response to these parameters, thereby providing insight into how environmental stress ornutritional status inuences rhizobial survival. Nutrients such as sucrose, phosphate and calcium enhance biolm formation as their concentrationsincrease, whereas extreme temperatures and pH negatively affect biolm formation.© 2006 Elsevier Masson SAS. All rights reserved.

Keywords: Biolm; Sinorhizobium meliloti ; Environmental signals

1. Introduction

The symbiosisbetween Gram-negative bacteria of the familyRhizobiaceae and the roots of legume plants (family Fabaceae)leads to the formation of nitrogen-xing nodules, in which thedifferentiated bacteria (bacteroids) reduce atmospheric nitrogento ammonia.

The study of root colonization, and subsequent nodule de-velopment and occupancy, has contributed a great deal to ourunderstanding of the nitrogen-xing association between rhi-zobia and legumes. Nevertheless, much is still unknown aboutrhizobial attachment to roots and the importance of its role in

* Corresponding author. E-mail address: [email protected] (W. Giordano).

subsequent stages of nodulation and nitrogen xation. Manymethods have been employed to study rhizobial attachment, butmost are indirect, involving initial detachment of the microor-ganisms from the root surface and then counting of them. Otherindirect procedures, such as radiolabelling bacteria, enzyme-linked immunosorbent assays or other biological assays, esti-mate the number of attached organisms in situ by measuringsome attribute of the attached organism [2]. Methods involv-ing direct observation, such as light-, laser-scanning confocal-,transmission electron- and scanning electron microscopy, de-tect microbial attachment directly and have greatly improvedour understanding of root colonization [13].

In a previous publication, we utilized a modied mi-crotiter plate assay to study biolm formation by Sinorhizo-bium meliloti and Rhizobium leguminosarum bv. viciae [14].

0923-2508/$ – see front matter ©

2006 Elsevier Masson SAS. All rights reserved.doi:10.1016/j.resmic.2006.06.002

8/17/2019 1-s2.0-S092325080kjghkhj6001264-main

2/9

868 L. Rinaudi et al. / Research in Microbiology 157 (2006) 867–875

Biolms are surface-attached communities of bacteria con-tained within a self-produced extracellular polymeric matrix;they are composed either of a single species or, more oftenin nature, of multiple bacterial species [15,49]. Indeed, themajority of bacteria appear to form biolms, and this multi-cellular mode of growth predominates in nature, most likely asa protective mechanism against hostile environmental condi-tions [7,20]. Biolm behavior affords bacteria, and especiallynon-spore formers such as rhizobia, a number of survival ben-ets, since biolms can be established on both abiotic andbiotic surfaces, often under stressful conditions. Research ona number of different bacteria indicates that biolms existas a mass of microcolonies in a single layer or as three-dimensional structures with vertical and horizontal channelsallowing liquid ow and dispersion of nutrients and wastecomponents [2]. The involvement of extracellular exopolysac-charides, pili, agellae and quorum sensing signals in biolmformation has been revealed by molecular genetic analyses

[12,40]. Also, recent research suggests that the environmentalsignals regulating whether bacterial cells will initiate a biolmdiffer from one bacterial species to another, thereby allowingeach bacterial species to efciently colonize its preferred envi-ronment [46].

The plant pathogen Agrobacterium tumefaciens , which per-sists as a surface-associated population of cells, developsbiolms on both inert surfaces such as soil particles, and liv-ing plant tissues [43]. Similarly, we previously reported that R. leguminosarum bv. viciae and S. meliloti establish biolmson both roots and abiotic surfaces [13,14]. Other studies havedemonstrated that bradyrhizobia and azorhizobia form biolmson fungal mycelia [44]. Although only a limited amount of information is known thus far about biolm formation in rhizo-bial species, it is tempting to speculate that biolm formation isimportant for the overall tness of rhizobia in the soil and in rhi-zosphere microenvironments, thereby contributing to efcientsymbiosis [54].

In the present report, we have expanded our analysis to studyS. meliloti biolm formation under various conditions of en-vironmental stress and nutrient status. These experiments lendsupport to our hypothesis that rhizobial biolm formation is im-portant for the survival of these non-spore forming bacteria insoil in the absence of a legume host.

2. Materials and methods

2.1. Bacterial strain, culture media and growth conditions

S. meliloti Rm1021 [29] was used in this study and wasgrown in minimal medium RDM ( Rhizobium Dened Medium)[51] supplemented with streptomycin (100 µg ml − 1) at 28 ◦ C.For strain maintenance, the medium was solidied with 1.5%Bacto-agar (Difco Laboratories). Bacterial liquid cultures com-prising 10–15% of the ask volume were grown in a NewBrunswick shaker at 200 rpm to mid- or late-log phase, dilutedto an optical density at 600 nm (OD 600) of approximately 0.2,and used in the biolm assay [14]. Bacterial growth and biolmformation were measured to determine their relationship under

the different conditions. RDM was supplemented with varioussugars, osmotic agents or salts as indicated in the Results sec-tion. When RDM was made with low phosphate levels (RDMnormally has a phosphate concentration of 12.5 mM), the pHwas adjusted with 2 M Tris–HCl, pH 6.8.

2.2. Biolm formation assay

2.2.1. Microtiter plate method The biolm formation assay used is based on the method

of O’Toole and Kolter [38] with modications [14]. This as-say relies on the ability of the cells to adhere to the wells of 96-well microtiter dishes made of polyvinylchloride. We foundthat polystyrene plates gave similar results (data not shown).To each well, 150 µl from an overnight culture was added. Af-ter inoculation, the plates were covered with plastic to preventevaporation and incubated without agitation at 30 ◦ C for a mini-mum of 24 h or for the time indicated in each experiment. Then,the contents of each well were gently aspirated with an auto-matic hand pipette or a Pasteur pipette. The wells were washedthree times with 180 µl of sterile physiological saline solutionand the plates were vigorously shaken in each wash in order toremove all non-adherent bacteria. The plates were emptied, leftto dry, and stained for 15 min with 150 µl per well of 0.1% CV.They were then rinsed thoroughly and repeatedly with waterand scored for biolm formation.

2.2.2. Quantication of biolm formationBiolm formation was quantied by addition of 150 µl

of 95% ethanol to each CV-stained microtiter dish well, and

the absorbance of solubilized CV was determined with a Mi-croELISA Auto Reader at 560 nm (series 700 microplatereader, Cambridge Technology) or at 570 nm, where indicated,in a BioRad microtiter plate reader (Model no. 680). Alterna-tively, and in order to compare the results from the microtiterplate reader, CV was solubilized in 200 µl of 95% ethanol.The entire contents were then transferred to a 1.5 ml Eppen-dorf tube, the volume was brought to 1 ml with distilled waterand absorbance was determined at 560 nm in a spectropho-tometer (DU-640 Spectrophotometer, Beckman Instruments).Before the addition of CV, cells were homogenized manuallyby repeated pipetting of the contents in each well and bacter-ial growth was quantied by measuring absorbance at 600 nm.Bacterial growth and adherence measurements were performedin triplicate and repeated at least three times; values were thenaveraged.

2.3. Potassium assay

Cells from 1 ml portions of bacterial suspensions (OD 600 =0.8–1.0) of free-living and sessile bacteria were harvested bycentrifugation for 1 min, supernatants were removed and pel-lets were suspended in 1 ml of deionizedwater. Thesuspensionswere boiled for 1 min to release potassium from the cells, andthe potassium concentrations (mM K + / g protein) were deter-mined by ame photometry as previously described [50].

8/17/2019 1-s2.0-S092325080kjghkhj6001264-main

3/9

L. Rinaudi et al. / Research in Microbiology 157 (2006) 867–875 869

2.4. Protein quantication

The protein concentration was determined by the method of Bradford [5] using bovine serum albumin (Sigma) as the stan-dard.

2.5. Microscopy

Biolms were established on 1 cm round glass cover slips,following our previously published procedure for using plas-tic tabs [14]. For microscopy, substrates and their associatedbiolms were washed with copious amounts of RDM to removeplanktonic cells. The samples were placed on a depression slideand viewed by uorescence microscopy using a Zeiss Axiophotmicroscope. Images were taken with Kodak Ektachrome Tung-sten 160T slide lm and processed using Adobe Photoshop.

Biolm cells were stained with the LIVE/DEAD ® Bac-Light™ bacterial viability kit according to the manufacturer’s

instructions (Molecular Probes, Eugene, OR). Briey, biolmsgrown on glass cover slips were incubated in the staining solu-tion (RDM containing 5 µM SYTO9 dye and 30 µM propidiumiodide) for 15 min in the dark at room temperature. These wereexamined under uorescence microscopy as described above.

2.6. Statistical analysis

Experiments were conducted in a completely randomizeddesign and were repeated. Values presented are the means of re-peated experiments. Data were subjected to one-way ANOVAfollowed by comparison of multiple treatment levels with thecontrol using post hoc Fisher’s LSD test. All statistical analy-ses were performed with Infostat software version 1.0.

3. Results

We had previously found that S. meliloti biolm formationwas greater when the bacteria were grown in RDM rather thanin richer media such as LB or TY [14], indicating that a nu-tritionally limiting environment increases the transition fromplanktonic to a sessile mode of life, i.e., a biolm. This obser-vation suggested that biolm formation may represent a sur-vival strategy in a nutritionally limited environment becausesurface colonization would provide a number of advantages

such as increased capture of nutrients that may be absorbedto surfaces [55]. Because the nutrient content of the growthmedium has been found to regulate the development of biolmsby other organisms [38,55], we tested various nutrients as wellas environmental conditions for their effects on the ability of S. meliloti to form biolms in the wells of the microtiter platedishes.

Fig. 1 shows that maximal biolm formation in RDM wasobserved 72 h after incubation in the microtiter plate wells, fol-lowed by a marked decrease in the amount of biolm formation.One explanation for the decrease in biolm formation is that itis related to the decrease in growth at later time points. Alter-natively, cells may detach from the wells at these time points.Shirtliff et al. [47] have shown that Pseudomonas aeruginosa

Fig. 1. Quantication of bacterial growth and biolm formation of static cul-tures of S. meliloti following different times of incubation. All assays wereperformed in triplicate, and mean values and standard deviations are shown.

Values having different letters are signicantly different from each other ac-cording to Fisher’s LSD test ( P < 0.05).

cycles in between increased biolm biomass and a signicantdetachment phase.

RDM normally contains 0.5% (ca. 0.015 M) of sucrose. Asthe concentration of sucrose was increased to 0.3 M (approx-imately 10%), biolm formation was augmented, by almosttwofold, as assayed by CV staining. In contrast, the highest in-crease in rhizobial growth occurred at 0.06 M sucrose (Fig. 2A).At 0.6 M sucrose, both bacterial growth and biolm formationwere signicantly diminished.

To determine whether the increase in biolm formation at0.3 M sucrose was correlated with a concomitant increase inthe osmolarity of the medium, we assessed the intracellularpotassium content in free-living bacteria. Accumulation of K +

followed by glutamate is a primary response to osmotic up-shift, by counteracting the loss of water, in both Rhizobiumand Agrobacterium when these bacteria are grown in minimalsalt medium [32]. A high value (2.16 mmol K + / g of protein)was observed for cells grown in RDM containing 0.3 M su-crose compared to those grown in medium containing 0.015M sucrose (1.29 mmol K + / g of protein), indicating that bac-teria undergo osmotic adaptation in RDM containing 0.3 Msucrose. Intracellular K + concentrations were also measured

for attached and planktonic S. meliloti incubated in RDM con-taining 0.015 M sucrose. The K + content in free-living bacteriawas 1.29 mmol K + / g of protein, whereas a value of 1.88 mmolK+ / g of protein was obtained for attached bacteria. Similar re-sults had been found previously for E. coli : the intracellularcontent of potassium ions was higher in attached bacteria thanin free-living bacteria [42].

We next assessed whether osmolarity had an effect on theability of S. meliloti to form biolms, by using NaCl and D-sorbitol as osmotic agents. Across a range of NaCl concentra-tions (0 to 0.3 M), Rm1021 growth was practically unaffectedexcept at 0.3 M NaCl, where growth was reduced (Fig. 2B).This is not surprising, as S. meliloti is reported to be toler-ant to 0.3 to 0.7 M NaCl (see references in [56]). In contrast,

8/17/2019 1-s2.0-S092325080kjghkhj6001264-main

4/9


Fig. 2. Bacterial growth and biolm formation of S. meliloti in RDM sup-plemented with (A) sucrose at the indicated concentrations and (B) differentlevels of NaCl and (C) D-sorbitol. All assays were performed in triplicate, andmean values and standard deviations are shown. Values having different let-ters are signicantly different from each other according to Fisher’s LSD test(P < 0.05).

biolm formation signicantly decreased at higher concentra-tions of NaCl (Fig. 2B). The inhibitory effect of NaCl on rhizo-bial biolm formation could be due to an osmotic effect, sincebiolm formation was highest in 0.3 M sucrose (equivalent tothe osmotic potential of 0.15 M NaCl) (Fig. 1A). Alternatively,

Fig. 3. Bacterial growth and biolm formation of S. meliloti in RDM supple-mented with (A) KNO 3 at the indicated concentrations and (B) different levelsof phosphate. The biolm was allowed to develop for 48 h before quantica-tion. All assays were performed in triplicate, and mean values and standarddeviations are shown. Values having different letters are signicantly differentfrom each other according to Fisher’s LSD test ( P < 0.05).

the inhibitory effect of NaCl could be attributed to a specicion effect, as suggested by Elsheikh and Wood [10]. Hence,we tested the responses of S. meliloti to increasing concen-trations (0–0.6 M) of another osmolyte, sorbitol. Compared togrowth, biolm formation was reduced in RDM containing 0.3and 0.6 M sorbitol (Fig. 2C). Taken together, these observationssuggest that both NaCl and sorbitol negatively affect biolm

formation through an osmotic effect, whereas the effect of su-crose on biolm formation is most likely nutritional and notrelated to osmolarity.

Growth and biolm formation were assessed using the mi-crotiter plate assay in RDM (normally 6 mM KNO 3) withdifferent nitrate concentrations (0–60 mM) (Fig. 3A). With in-creasing nitrate concentrations, S. meliloti growth increased,reaching a plateau at about 30 mM. Biolm formation was lowbetween 0 and 3 mM and increased signicantly starting at6 mM of nitrate, with a maximum at 30 mM. At 60 mM ni-trate, biolm formation signicantly decreased in contrast torhizobial growth, which remained at a high level (Fig. 3A).

Apart from carbon, nitrogen and oxygen, phosphorus is themost important nutrient for living cells. To test the importance

8/17/2019 1-s2.0-S092325080kjghkhj6001264-main

5/9


Fig. 4. Quantication of bacterial growth and biolm formation of static cul-

tures of S. meliloti (A) at different CaCl 2 and (B) MgSO 4 concentrations. Allassays were performed in triplicate, and mean values and standard deviationsare shown. Values having different letters are signicantly different from eachother according to Fisher’s LSD test ( P < 0.05).

of phosphate in biolm formation, S. meliloti cells were grownin microtiter plate wells in RDM with phosphate levels rangingfrom 0 to 25 mM. At low phosphate concentrations (lower thanthe normal concentration in RDM, 12.5 mM), weak biolmformation was observed compared to the higher phosphate con-centrations (Fig. 3B). This increase in biolm establishmentcorrelated with an increase in rhizobial growth.

We next tested the involvement of Ca 2+ and Mg2+ in

biolm formation. The optimal Ca2+

concentration for growthwas 0.7 mM, which is the concentration of Ca 2+ in RDM.Although increasing the Ca 2+ concentration above this op-timum had a slight deleterious effect on growth, it had apositive effect on biolm formation (Fig. 4A). The same in-crease in biolm formation was observed with increasing Mg 2+

concentrations. Concentrations above 1 mM of Mg 2+ (theconcentration of Mg 2+ in RDM) had little effect on growth(Fig. 4B).

S. meliloti strains are extremely sensitive to acidic pH andwill show good growth only above pH 5.5 [17,18]. RDM, likemost rhizobial media has a pH close to 7.0, and both biolmformation and growth were high at this pH. In contrast, at pH4.0, which is typically found in acidic soils, bacteria attached

Fig. 5. (A) Microtiter plate assay of biolm formation by Rm1021 grown for24 h before staining with crystal violet. Each value point is the average of atleast 16 wells. Error bars indicate the standard deviation from the mean. Thebiolm levels were standardized to the different growth rates at each pH. (B, C)Fluorescent microscopy of biolms grown on glass coverslips under differentpH conditions for 6 d. Biolms were stained with LIVE/DEAD uorescentmarkers. Bar = 10 µm. (B) Cells grown at pH 7.0. (C) Cells grown at pH 4.0.

to the microtiter plate wells in numbers that were even greaterthan those observed at neutral pH (Fig. 5A). This was a surpris-ing result, so we stained the biolm cells with the LIVE/DEADstain to determine cell viability. At neutral pH, the biolm tow-

8/17/2019 1-s2.0-S092325080kjghkhj6001264-main

6/9


Fig. 6. Inuence of pH on growth and biolm formation by strain Rm1021grown for 48 h in RDM at different pH values (6.0–8.0). All assays were per-formed in triplicate, and mean values and standard deviations are shown. Valueshaving different letters are signicantly different from each other according toFisher’s LSD test ( P < 0.05).

Fig. 7. Bacterialgrowth andbiolm formationof S. meliloti under different tem-perature regimes. Bacteria were grown at 22–45 ◦ C. All assays were performedin triplicate, and mean values and standard deviations are shown. Values havingdifferent letters are signicantly different from each other according to Fisher’sLSD test (P < 0.05).

ers uoresced green after staining, indicating that the cells wereviable (Fig. 5B); a few cells in the older parts of the biolm u-oresced red (data not shown). At pH 4.0, nearly all cells werein densely packed mats that stained red rather than in towers(Fig. 5C). However, these cells were not necessarily dead. Vir-tually all planktonic cells also uoresced red at this pH, but theywere observed to swim and twitch, indicating that the bacteriawere viable. We also analyzed biolm formation at initial pHvalues (from 6.0 to 8.0) that were less likely to inhibit rhizobialgrowth (Fig. 6). As expected, biolm formation was signi-cantly diminished at the extremes, pH 6.0 and 8.0, whereas thegrowth of S. meliloti cells, as measured by absorbance at OD 600 ,was not. Taken together, these data show that the optimal pH forgrowth is pH 7.0 (Fig. 6). However, at pH 4.0, although growthwas low, the number of cells in the biolm was very high.

We also studied the inuence of temperatures ranging be-tween 22 and 45 ◦ C on growth and the ability of S. melilotito form biolms (Fig. 7). At 28 or 37 ◦ C, no clear differencein growth or biolm accumulation on microtiter plate wellswas observed, but the temperatures at the extremes negativelyaffected both bacterial growth and biolm formation. Interest-ingly, at 45 ◦ C, biolm formation was observed at levels com-parable to 22 ◦ C.

4. Discussion

Bacterial biolms have a signicant impact in medical, in-dustrial and environmental settings. Moreover, a number of environmental parameters inuence whether biolms are suc-cessfully established in these settings. For example, osmolarityclearly inuences rhizobial biolm formation, as was foundfor P. uorescens [38] and E. coli [41]. We found that growthat high osmolarity (and not simply ionic strength) inhibited

biolm formation in S. meliloti. S. meliloti cells within biolmsare likely to encounter higher osmolarity conditions. Indeed,the intracellular concentration of potassium ions, which is es-sentially proportional to the osmolarity of the external mediumin the absence of exogenous solutes such as proline or betaine,is almost 1.5-fold higher in attached bacteria than in free-livingbacteria (see results). Taken together, these results show that S.meliloti cells within the biolm are under more osmotic stressthan planktonic cells, and also that high osmotic potential in-hibits biolm establishment. Furthermore, the effects of sucroseon S. meliloti biolm formation are not strictly related to osmo-larity, but rather to the effects of this carbon source on growth,at least until very high osmotic potentials were reached, andthen growth was inhibited.

Although we had previously found that nitrogen was re-quired for proper biolm formation [14], we did not determinethe optimal concentration of nitrate. In this report, we foundthat for optimal biolm formation a threshold of nitrate con-centration between 3 to 6 mM is required, below which thereis minimal biolm formation and above which biolm forma-tion is at a maximum. The increase in biolm establishment wasunrelated to growth, since the biolm/growth ratio in minimalmedium containing 6 mM nitrate was close to 1.0, whereas in3 mM nitrate, the ratio was ca. 0.22. This means that in RDMwith 6 mM nitrate, that ratio was almost vefold higher than

in RDM containing 3 mM nitrate. This result is interesting inlight of the fact that the addition of as little as 5 mM nitrateto seedling growth medium is reported to signicantly decreasethe number of rhizobial cells adhering to seedling roots [56].Indeed, a number of studies have illustrated the negative effectof nitrate on root infection by rhizobia [48]. However, differ-ences in tolerance to nitrate and ammonium have also beenfound among rhizobial isolates when investigated in nodulationsystems [34]. For example, Gibson and Harper [16] reportedthat different strains of B. japonicum have varying toleranceto external nitrogen application in their nodulation and nitro-gen xation characteristics. The inhibitory effect of nitrogen onrhizobial cell adherence to roots and on nodulation may be inpart plant-mediated because excess nitrate is known to inuence

8/17/2019 1-s2.0-S092325080kjghkhj6001264-main

7/9


lectin activity [9], and legume lectins mediate a specic typeof adherence of rhizobia in Rhizobium -legume symbiosis [22].However, non-lectin-mediated bacterial attachment also occurs,and an excess of nitrate that inhibits the symbiotic process maynot necessarily affect non-specic binding to abiotic surfaces.This would explain why adherence to microtiter plate wells isnot inhibited by concentrations of nitrate that inhibit nodula-tion. Therefore, our observations are probably more relevant tothe ability of rhizobia to adhere to different soil surfaces.

Availability of phosphate is critical for bacterial growth andmetabolism. To some extent, the increase in S. meliloti growthcorrelated with augmented surface attachment as the phosphateconcentration was increased. It is well recognized that rhizo-bia have developed physiological mechanisms for coping withphosphate starvation [26] because the concentration of freephosphate is usually low in natural environments, such as soiland water. For example, in S. meliloti Rm1021, exopolysac-charide biosynthesis is inuenced by phosphate starvation, re-

sulting in decreased production of succinoglycan (EPSI) andinduction of galactoglucan (EPSII) biosynthesis [30,57]. Thelarge difference in phosphate concentrations between the soil(typically 1 to 10 µM) [3] and the nodule (up to 100 mM) [25]is an excellent example of the strikingly dissimilar conditionsthat Sinorhizobium encounters during its lifetime, requiring cer-tain physiological adjustments to occur in order to survive inthese environments. For example, the phosphate concentrationmay serve as a signal to regulate which EPS is produced by S.meliloti according to cues taken from the surrounding environ-ment [30]. The type of EPS produced may have an importanteffect on biolm formation because EPS accounts for 50–90%of the total organic carbon of biolms [11]. We previouslyshowed that EPS-minus S. meliloti mutants are compromisedin biolm formation, whereas Exo-plus mutants make largerbiolms than wild-type S. meliloti [14]. Although EPS variesin chemical and physical properties, it is primarily composedof polysaccharides. Some of these polysaccharides are neutralor polyanionic, as is the case for the EPS of Gram-negativebacteria. The presence of uronic acids (such as D-glucuronic, D-galacturonic, and mannuronic acids) or ketal-linked pyruvatesconfers anionic properties to the EPS [49]. This is importantbecause it allows the association of divalent cations, such asCa2+ and Mg 2+ , which have been shown to cross-link withEPS, thereby providing greater stability in a biolm [11].

In bacteria, Ca2+

is implicated not only in EPS binding,but also in processes as important and diverse as the cell cy-cle and cell division, competence, pathogenesis, motility andchemotaxis, and quorum sensing [23,31,35,36,45]. In the ex-tracellular space, Ca 2+ also has an important structural role,maintaining the integrity of the outer lipopolysaccharide layerand the cell wall [45]. It is well known that the attachment abil-ity of the bacteria to roots varies considerably depending on thebacterial strain, nutrient requirements and growth conditions.Caetano-Anollés et al. [6] suggested that Ca 2+ could act as abridge between negatively charged groups on plant and bacter-ial surfaces, and/or indirectly activate the bacteria for adhesion.When the medium was supplemented with Ca 2+ ranging from7 to 28 mM, a clear increase in the rate of accumulation of at-

tached bacteria occurred in the microtiter plate wells. This mayresult from the interaction between EPS and Ca 2+ . However,higher levels of Ca 2+ negatively affected rhizobial growth, sowe examined the effect of another divalent cation (Mg 2+ ) onbiolm formation, and found that high Mg 2+ concentrations ledto an increase in biolm formation, but without an inhibitoryeffect on growth. The exact mechanism whereby the transitionto the mature biolm state occurs is unknown, but a decreasein growth may incite the bacteria to form a biolm or, alter-natively, the presence of high levels of divalent cations maystabilize EPS in the biolm.

The optimal pH for S. meliloti growth formation was pH 7.0,but more bacteria attached to the microtiter plate well at pH4.0. As observed with Streptococcus gordonii [27], S. melilotibiolm formation is more sensitive to pH changes than bacte-rial growth. However, pH had no effect on biolm formationin P. uorescens [38]. Taken together, these results suggest thatpH effects, in terms of establishing a biolm, may differ from

one bacterial species to another, thereby enabling each bacterialspecies to efciently colonize its preferred environment.

Differences in tolerance of high temperatures among speciesand strains of Rhizobium have long been recognized [4,19,24,28,37,39,52]. For most rhizobia, the optimum temperaturerange for growth in culture is 28–31 ◦ C, and many are unableto grow at 37 ◦ C [18], although Allen and Allen [1] noted thatS. meliloti has an optimum growth temperature of 35 ◦ C, andeight of eleven strains tested by Graham and Parker [19] grew at39 ◦ C. We found that S. meliloti grew in a wide range of temper-atures and no differences were observed between 28 and 37 ◦ C.Moreover, biolm formation was also affected by extreme tem-

peratures, in that more attachment was evident at 28 and 37◦

C.The high tolerance of Rm1021 towards elevated temperaturesmay allow this strain to survive periods of thermal stress in thesoil. For example, high (not extreme) soil temperatures delaynodulation or restrict it to the subsurface region [18]. Munns etal. [33] found that alfalfa plants grown in desert environmentsin California maintained few nodules in the top 5 cm of soil, butwere extensively nodulated below this depth. Under these con-ditions, nitrogen xation activity is shut off, but may resumewhen temperatures are lower.

The results of the present study show that external condi-tions exert an important inuence on the partitioning of cells

between planktonic and biolm phases. At this point, it isnot clear whether this partitioning is the simple result of ashift in physiology, possibly related to quorum-sensing cues,which are important for biolm development in P. aeruginosa[8] and Mesorhizobium huakuii [54], or whether there is anunderlying genetic switch, such as that seen with phase vari-ation by some species facilitating adaptation to new ecolog-ical conditions [21], with the subsequent selection and out-growth of adherent-phase variants in a time- and population-size-dependent manner. The former seems more likely, sinceseveral microarray studies comparing biolm to planktonic bac-teria have shown that there are signicant differences in geneexpression between these two states of existence [53], implyingthat there is a concomitant change in physiology.

8/17/2019 1-s2.0-S092325080kjghkhj6001264-main

8/9

8/17/2019 1-s2.0-S092325080kjghkhj6001264-main

9/9


and biolm formation in Escherichia coli via regulation of the csgD gene,J. Bacteriol. 183 (2001) 7213–7223.

[42] C. Prigent-Combaret, O. Vidal, C. Dorel, P. Lejeune, Abiotic surface sens-ing and biolm-dependent regulation of gene expression in Escherichiacoli , J. Bacteriol. 181 (1999) 5993–6002.

[43] B.E. Ramey, A.G. Matthysse, C. Fuqua, The FNR-type transcriptional reg-ulator SinR controls maturation of Agrobacterium tumefaciens biolms,Mol. Microbiol. 52 (2004) 1495–1511.

[44] G. Seneviratne, H.S. Jayasinghearachchi, Mycelial colonization bybradyrhizobia and azorhizobia, J. Biosci. 28 (2003) 243–247.

[45] R.J. Smith, Calcium and bacteria, Adv. Microb. Physiol. 37 (1995) 83–133.

[46] N.R. Stanley, B.A. Lazazzera, Environmental signals and regulatory path-ways that inuence biolm formation, Mol. Microbiol. 52 (2004) 917–924.

[47] M.E. Shirtliff, J.T. Mader, A.K. Camper, Molecular interactions inbiolms, Chem. Biol. 9 (2002) 1–20.

[48] J. Streeter, Inhibition of legume nodule formation and N 2 xation by ni-trate, Crit. Rev. Plant Sci. 7 (1988) 1–23.

[49] I.W. Sutherland, Biolm exopolysaccharides: A strong and sticky frame-work, Microbiology 147 (2001) 3–9.

[50] L. Sutherland, J. Cairney, M.J.I. Elmore, R. Booth, C.F. Higgins, Osmoticregulation of transcription: Induction of the proU betaine transport gene

is dependent on accumulation of intracellular potassium, J. Bacteriol. 168(1986) 805–814.

[51] J.M. Vincent, A Manual for the Practical Study of Root-Nodule Bacteria,Blackwell Scientic Publications, London, 1970.

[52] J.M. Vincent, Rhizobium : General microbiology, in: R.W.F. Hardy, W.S.Silver (Eds.), A Treatise on Nitrogen Fixation, Section III, Biology, Wiley,New York, 1977, pp. 277–366.

[53] R.D. Waite, A. Papakonstantinopoulou, E. Littler, M.A. Curtis, Transcrip-tome analysis of Pseudomonas aeruginosa growth: Comparison of geneexpression in planktonic cultures and developing and mature biolms,J. Bacteriol. 187 (2005) 6571–6576.

[54] H. Wang, Z. Zhong, T. Cai, S. Li, J. Zhu, Heterologous overexpression of quorum-sensing regulators to study cell-density-dependent phenotypes ina symbiotic plant bacterium Mesorhizobium huakuii , Arch. Microbiol. 182(2004) 520–525.

[55] J.W.T. Wimpenny, R.A. Colasanti, A unifying hypothesis for the structureof microbial lms based on cellular automation models, FEMS Microbiol.Ecol. 22 (1997) 1–16.

[56] H.H. Zahran, Rhizobium -legume symbiosis and nitrogen xation undersevere conditions and in an arid climate, Microbiol. Mol. Biol. Rev. 63(1999) 968–989.

[57] H.J. Zhan, C.C. Lee, J.A. Leigh, Induction of the second exopolysaccha-ride (EPSb) in Rhizobium meliloti SU47 by low phosphate concentrations,J. Bacteriol. 173 (1991) 7391–7394.

Documents

1-s2.0-S092325080kjghkhj6001264-main